Using a novel automated approach for quickly building and testing thousands of battery cells, Wildcat Discovery Technologies, a startup in San Diego, California, has developed new materials that could boost the storage capacity of lithium-ion batteries for cars and portable electronics by more than 25 percent.
Batteries based on the new materials could extend the range of electric vehicles or allow automakers to keep the same range but use fewer battery cells, thus reducing the cost of the battery pack, an electric car's most expensive part. Work still needs to be done to improve the durability of the new materials, but the results provide validation for Wildcat's high-throughput screening technique that allows researchers to quickly sort through combinations of materials.
High-throughput screening is common in the drug and chemical industries to discover new compounds and catalysts, and the technique has made inroads into battery development. What makes Wildcat's process different is that it makes complete battery cells, not just individual parts of the cells, such as the electrodes. This is important because the performance of any given material in the cell depends on how it interacts with the other parts. "With conventional approaches, you get a lot of false positives and false negatives," says Steven Kaye, Wildcat's chief scientific officer. An electrode material that looks promising on its own may fail in a complete cell as it interacts with electrolytes, additives, and the opposite electrode, he says. And one that looks mediocre could improve markedly when mixed with other materials in a cell.
Wildcat's new material, a variant of lithium cobalt phosphate, is one that would ordinarily have been rejected because it operates at a voltage that quickly destroys the battery's electrolyte, the liquid that conducts lithium ions between the electrodes. But the researchers paired the material with many new electrolyte recipes and eventually discovered one that could survive the high voltage. In all, the company screened 4,000 materials over a period of about four months to find the ones that worked.